Treat me as a complete newb; I experimenting/researching and have a raw recorded kick drum. Im investigating just removing <40hz and thought you would simply draw the freq range and hit delete…right? but there is only a very slight effect
What am I doing wrong?
Yeah, well, SL doesn’t really do hard cuts IME
I find I often have to re-select multiple times and cut the exact selection 3-5 times…sometimes many more
I don’t know why, that’s just my experience
(default key command is CTRL+SHIFT+D for reselect)
I find that non-precise selection habits typically sound better than attempting surgical precision. It’s more like you are working on medium congealed jelly than cutting wood. SL just wants you to massage transforms…I dunno if that’s just the way it is with spectral editing, but is the experience I’ve gleaned over the past 2.5 years. It’s like moulding clay that is too wet…take something away and the sounds freqs just ooze back into where you just scooped out
if using the Freq Range tool, even setting the frequency fade to 0.00oct isn’t going to yield a hard cut
I understand the windowing etc but for experimental stuff…would be nice if it didnt just reconstruct but allowed sound design limits. In the last couple of mins I worked out it must be freq related because the higher the window, the more it behaves as expected.
Even when I adjust the windowing, the blur of transient is pretty similar to Linear Phase although it doesnt seem symmetric.
I would have thought eg applying a gain module to that region would work ie to attenuate but that does not work either
That’s one of those bugs, imho.
When hitting “DEL”, delete is meant, nothing else.
To clarify: If delete needs several steps, i.e. pressing “DEL” several times to accomplish the delete, than this should be possible to do this in one step also. No “if’s and but’s”, not some sophisticated “workaround” found out by some user - just a simple plain “DEL” must suffice.
Okay! Quick correction:
It’s not a bug/glitch and actually works as intended. The best way to explain this is using the concept/scenario of “linear” vs “non-linear" eq and how it works. A “linear" eq boosts and cuts frequencies however it is never an exact cut/boost because there are random errors around the sharp corners of those boosts/cuts (basically it will always introduce phasing). A “non-linear" eq (Believe-it-or-not) actually works just like a linear eq, however the boosts/cuts is designed to be as less noticeable around the edges as possible (meaning it can do sharp cuts and even more sharper/precise boosts but it will also interfere with the phase, it’s just less noticeable).
Yep I get that…did a lot of research of pros/cons of linear phase on sub freq…actually the best results (so far) in that region are oversampling with a mixed phase for min impact. Thats why TDR use it on infrasonic. For anyone who wants a great tool for free…ozone eq can do the mix in infinite variability and is a great EQ
BUT
I was looking for a way to do offline thinking that SL should be able to do both ie phase and temporal but it seems it still brings pre ring (window smoothing?). The only way delete works down there is max window, otherwise the resolution just cant “see” it. I did a bunch of research that backed it up but even with large windows it doesnt actually “work” that well so Im back to square one…just use mixed phase but putting them all in blind tests…all have cons. Im just trying to tweak with better mic placement which has gotten rid of the sub 40hz a lot, the fundamental is around 50 but I tune for diff songs, its not EDM but just transients are nicer but you can of course work it the other way…just so context dependent.
From what I can see SL cannot do what Im asking
@Aurasphere Try the following:
This will give you the result you’re looking for.
Note
It may sometimes be important to stick to the above steps in the correct order to obtain the desired result. The delete action when using the Frequency Range Selection tool is dependent upon both the Frequency Fade setting and the chosen FFT size. Larger FFT size settings will give you more severe frequency deletions, if that’s what you need.
Wow! That does work! FFT size absolutely critical for this =-D
Cheers @stingray
Yep I finally got to that stage and worked it out…but appreciate
However, if you do some forensics and isolated listening, the result is pretty comparable to Linear Phase in regards to pre ring. Freq iin that region just dont need smear like that especially with long wavelengths. I have a particular and strict philosophy as its taken quite a couple of years to get there…but it looks like its not really physically achievable without serious compromise.
The best it gets seems to be TDR Infrasonic or very fine control with mixed mode in eg Ozone eq
If I get time tomoz Ill post some screen shots of overlaid results. With a couple of kws of subs pumping this…ie FOH, its def noticeable…Im not dreaming hehe. I also have a tactile magnifying glass that confirms it.
Great job.
Still waiting for a version of SL, where each and every function/method is marked as dependend of FFT-size and/or windowing.
It could be done in another textcolor or a ToolTip.
Yep…I asked for that too…UX is so important
I would add a coloured symbol and tooltip with recommended start min/max